This appeared in the Sunday 27 Aug 00 edtion of the San Jose Mercury News.

I don't know Mr. Cheriton, yet I do think he makes some interesting points.

-Mike Cheponis

p.s. I am -not- saying "Rip IPv6 out of NetBSD" or anything of the sort.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Thumbs-down on IPv6

Regarding Dan Gillmor's Aug. 20 column on Internet Protocol Version 6 (IPv6),
I can reasonably claim to be a ``networking specialist'' (as someone who has
been teaching and researching networking at Stanford for the past 18 years),
and I do not give IPv6 a thumbs-up. In fact, to me it is clear that IPv6 is
dead on arrival, or more accurately before arrival, given that it hasn't
arrived. The IPv6 ``solution'' is somewhat like replacing English with
Esperanto to make spelling easier. Maybe it would, but it's not going to
happen. The compelling reason to make this drastic change is just not there.

Most of the fantastical benefits its promoters claim of security, quality of
service, connectivity, etc., do not stand up to technical scrutiny, and what
little does is totally uncompelling in the face of the massive cost of
getting it deployed.

The IPv6 effort was started in 1992, before the advent of the Web, before
Bill Clinton's presidency, before Netscape, before dot-coms and even before
the commercializing of the Internet, making it prehistoric in Internet
terms. It grew out of a concern that the Internet was running out of
addresses. However, that problem was effectively solved by the introduction
of what is called network address translation (NAT). Since 1992, NAT has been
deployed widely in the Internet. By any market measure, IPv6 completely lost
out to NAT.


David Cheriton

Computer Science Department
Stanford University